multi-group SEM, CFA, and Path Analysis
Posted: Tue Aug 11, 2026 5:53 pm
Hello All,
First of all, thank you for this wonderful program! It’s going to help me tremendously in working with grad students.
I’ve run into a question regarding multi-group analysis. Namely, I want to compare two simple path models between husbands and wives (this isn’t paired data btw). This is using synthetic data for purposes of illustration. The basic model is:
CompulsiveInternetuse -> Intimacy -> Commitment
The grouping variable is huswiv.
I first tried to run it via the SEM syntax but I get an error message:
undefined columns selected
Debug
Error in `[.data.frame`(data, , var): undefined columns selected
private$.init()
Datamatic$new(self)
initialize(...)
private$.inspect_data()
levels(data[, var])
data[, var]
`[.data.frame`(data, , var)
stop("undefined columns selected")
However, when I run it as a path analysis, everything runs fine.
I’m including a omv file with both the path and the SEM analysis here, but I don’t have the grouping variable in place for the SEM version because it throws an error. Again, the grouping variable is “huswiv”. I’m hoping someone can help me understand this problem.
Thank you in advance!
Yours,
Mary
First of all, thank you for this wonderful program! It’s going to help me tremendously in working with grad students.
I’ve run into a question regarding multi-group analysis. Namely, I want to compare two simple path models between husbands and wives (this isn’t paired data btw). This is using synthetic data for purposes of illustration. The basic model is:
CompulsiveInternetuse -> Intimacy -> Commitment
The grouping variable is huswiv.
I first tried to run it via the SEM syntax but I get an error message:
undefined columns selected
Debug
Error in `[.data.frame`(data, , var): undefined columns selected
private$.init()
Datamatic$new(self)
initialize(...)
private$.inspect_data()
levels(data[, var])
data[, var]
`[.data.frame`(data, , var)
stop("undefined columns selected")
However, when I run it as a path analysis, everything runs fine.
I’m including a omv file with both the path and the SEM analysis here, but I don’t have the grouping variable in place for the SEM version because it throws an error. Again, the grouping variable is “huswiv”. I’m hoping someone can help me understand this problem.
Thank you in advance!
Yours,
Mary